Python · Rust · macOS / Windows / Linux

Audio for Python, without ffmpeg.

sound_rs plays and decodes the formats people actually have — MP3, FLAC, WAV, OGG, M4A — from a single wheel. The decoders are compiled in, so there is nothing to install besides the package itself.

pip install sound_rs

import sound_rs

with sound_rs.Player() as player:
    player.play_file("music.flac")   # streams from disk, returns immediately
    player.seek(42.0)
    print(player.position, player.duration)
    player.wait()

What it does

• Playback of files and NumPy arrays, blocking or through a persistent player with pause, seek, volume and position.
• Streaming decode: a two-hour file never sits in RAM, and the first sound arrives in milliseconds instead of after a full decode.
• Decoding to arrays for analysis, one file at a time or in parallel.
• Metadata straight from the container header, without decoding the audio.
• Typed: ships sound_rs.pyi and py.typed, so editors autocomplete every call.

What it does not do

No effects, no mixing, no recording yet, and no plugin system. If you need a DAW-style graph, this is the wrong tool; if you need to play or load audio reliably, it is the right one. The benchmark page also says where it is slower than the alternatives.

Where to go next

• Installation — platforms, wheels, Linux audio libraries.
• Quickstart — the five things most code does.
• Command line — soundrs --help without writing Python.
• The player — controls, seeking, lifecycle.
• API reference — every function and method.